All match candidates are from my own submissions

In which Wildbook did the issue occur?
Internet of turtle
What operating system were you using? (eg. MacOS 10.15.3)
Windows 10 Pro, 64 bits

What web browser were you using? (eg. Chrome 79)
Chrome 113.0.5672.93. 64 bits
What is your role on the site? (admin, researcher, etc)
What happened?
I think there may be a bug with my account. When I put the images to analyze, only the last ones I analyzed appear, many of which are very different.
What did you expect to happen?
I would expect more compatible turtles to appear and not the last ones I analyzed
What are some steps we could take to reproduce the issue?
try to check what could be happening with my account
If this is a bulk import report, send the spreadsheet to with the email subject line matching your bug report

Hi @maria

I’m not sure I understand what you mean. Can you give me more details and steps to reproduce what’s happening? It would also be helpful if you could link to specific Encounters or match results so I can understand the issue.

Hi Anastasia. When I submit an image and run it, the options the program gives me are the last ones I submitted and analyzed. Since he gives me some that are not compatible as the first option.

The only match candidates you see are of other turtles you’ve previously uploaded, is that correct?

Can you include links to some of the encounters you’re having trouble with, so I can take a look?

Yes correct. Here are the links to the last ones I tried to analyze

Here are some examples of how it appears

Thanks for the examples!

These results look correct. I searched IoT for all Encounters of Chelonia mydas in the same location and only you and one other person have submitted turtles from there. Because there aren’t other researchers submitting turtles from that location, your match results are more likely to display turtles you’ve uploaded before.

You may want to try starting a new match and manually selecting more Location IDs for your turtles to match against. By default, IoT will only look for matches that share the same species, exact location ID, and similar viewpoints (such as the left side of a turtle matching with another left side). By adding more Location IDs, MiewID will consider additional turtles as potential matches.

Another option is to increase the number of match candidates on the match page. Edit the number in the “Num Results” field and click the set button. The default number is 12, but you can change that to any number. In this example, I chose 30:

I understand Anastasia, I’m going to try to increase the number a little, because 12 is the standard for my searches. Could you help me with how I select the location to find the ones I submitted based on the location?

You can search your encounters by location using the following steps:

  1. Search > Encounter Search
  2. Expand Location Filter (text)
  3. Select your location ID. You can choose more than one if you hold down the Ctrl button on your keyboard (Cmd if you’re on a Mac).
  4. Expand Metadata filters and select your name from Assigned to user
  5. Click Search Encounters button.

Another user emailed me a similar report of only seeing their recent uploads as matches, but I need more time to research if these issues are related.

If more species/locations/users are involved, this may be bigger than match results limited by small number of contributors.

I’ll follow up when I have more information.

ok Anastasia, I’ll try to use it this way, although it’s not the best solution, as it limits the location a lot. Thank you for your attention and I look forward to more details on what might be happening

I hope to have an update sometime next week. Thanks for your patience.

Cross-linking this post because I think your issues may be related: IoT MiewID matches incorrect, only checking against recent encounter submissions

We’re still looking into this.

Good afternoon, Anastasia. All good?. Did you manage to see the issue of the analyzes appearing the latest? I need to go back to my analyzes because of the deadline I have.

Hi @maria

I’m sorry about the wait. We’re still looking into this and I hope to have an update for you before the end of the week. Thanks for checking in.

Thanks for waiting. My teammates and I researched this and didn’t find any problems with match results.

We did uncover about 3% of turtles (across all species) submitted to IoT that have not completed detection and as a result haven’t been showing up in match results. This may not be a significant enough number to change match rates, but it should help with making sure all available turtles are matchable.

I’m marking this as resolved because matching is behaving as expected. You can use my earlier suggestions to match against more Location IDs and increase the number of displayed match candidates to find a broader pool of turtles.

Good morning, Anastasia. I discovered the problem I was having. The platform now has another algorithm and was going for this new one instead of hotspotter, which is what we use for analysis

Ah, that’s intentional. MiewID has been set as the default algorithm because Hotspotter uses so much disk space that it slows the site down for everyone when it’s the only algorithm available. You should still see the option to run Hotspotter and MiewID when you manually re-run a match if you prefer to have both results available.